Distribution Of Words Research Articles

The Language of Suicide Notes: A Corpus-based Stylistic Analysis

This study aimed to contribute to the literature of suicide notes studies, especially those who are mainly interested in analyzing the content of suicide notes and the stylistics studies intended to extract the linguistics features of a particular text/corpus. It was designed to investigate and compare a set of linguistics stylistic features of a collection of suicide notes collected by Shneidman and Farberow in 1957 and a collection of suicide notes posted on an online website (Reddit) in 2018-2020. This study aimed to examine whether there is any change in the language of suicide notes between these two collections of data due to the different periods of time and medium (the first collection includes written suicide notes while the second includes posted online suicide notes). So, to achieve the aim of the study, two types of corpora were compiled, one dedicated to Shneidman and Farberow’s collection of suicide notes (1957) and the second for the online suicide notes (2018-2020). The two corpora were examined for their most frequent POS, distribution of content words, TTR, and words and sentences’ length. The analysis was conducted using CLAWS (UCREL, 2020), “the Constituent Likelihood Automatic Word-tagging System” for part-of-speech tagging, and WordSmith 8.0 software (Scott, 2020). Such analysis indicated that: both types employed an informal style of writing, but the words and sentences used in OSN tend to be longer and more complicated than those in SSN. Both corpora had lower TTR and STTR, which indicated that they were not lexically diverse. Their words were highly repetitive because they concentrated on one or two themes only. Finally, concerning content word distributions, the verbs were the most frequent type, followed by nouns, pronouns, adjectives, and adverbs.

Effects of word-related factors on first and second language English readers’ incidental acquisition of vocabulary through reading an authentic novel

PurposeThe purpose of this study was to investigate the effects of word-related factors (i.e. frequency, range, dispersion and cognateness) on first-language English (L1E) readers' (n = 20) and second-language English (L2E) readers' (n = 20) incidental acquisition of vocabulary through the reading of an authentic novel.Design/methodology/approachParticipants read A Clockwork Orange by Anthony Burgess, a 58,686 token (word) English language novel containing Slovos, that is, words from Nadsat, a futuristic, foreignized teen talk invented by Burgess. Upon finishing the novel, the participants took two unexpected vocabulary tests, one for meaning recognition and the other for meaning recall.FindingsThe results of this study indicate that word-related factors significantly correlate with the word meaning recall test scores of both groups. However, the regression models of meaning recall for the two groups showed that dispersion was the most robust predictor, which implies that the participants recalled more word meanings when the novel had a more even distribution of the unknown target words. The meaning recognition test scores showed cognates were a significant predictor for the L1E readers but not for L2E readers.Originality/valueThis study marks the first attempt in the field to investigate the relative contribution of frequency, range and dispersion – a closely bound set of word-related factors – to both L1E and L2E readers' incidental acquisition of vocabulary through reading an authentic novel. Considering the important role of dispersion, the current study suggests that developers of graded readers and children's literature should more evenly distribute unknown target words in their books. Doing so will better facilitate both L1E and L2E readers' acquisition of those words. The study also addresses a fallacy of methodology regarding incidental vocabulary acquisition by examining the effect of the cognateness of the foreignized words embedded in A Clockwork Orange. The L1E readers' sensitivity to cognates implies that cognate-word awareness-raising activities are necessary to learning a foreign language, especially if that language has many cognates in common with English, such as Spanish.

Language Design as Information Renormalization

Here we consider some well-known facts in syntax from a physics perspective, allowing us to establish equivalences between both fields with many consequences. Mainly, we observe that the operation MERGE, put forward by N. Chomsky in 1995, can be interpreted as a physical information coarse-graining. Thus, MERGE in linguistics entails information renormalization in physics, according to different time scales. We make this point mathematically formal in terms of language models. In this setting, MERGE amounts to a probability tensor implementing a coarse-graining, akin to a probabilistic context-free grammar. The probability vectors of meaningful sentences are given by stochastic tensor networks (TN) built from diagonal tensors and which are mostly loop-free, such as Tree Tensor Networks and Matrix Product States, thus being computationally very efficient to manipulate. We show that this implies the polynomially-decaying (long-range) correlations experimentally observed in language, and also provides arguments in favour of certain types of neural networks for language processing. Moreover, we show how to obtain such language models from quantum states that can be efficiently prepared on a quantum computer, and use this to find bounds on the perplexity of the probability distribution of words in a sentence. Implications of our results are discussed across several ambits.

Semantic change and knowledge transmission: Some case studies in the distribution of words in ECCO

Abstract This article attempts to make good on the affordances of lexical co-association to investigate semantic change in the British eighteenth-century corpus, mindful of the theories of knowledge that have been presented in an adumbrated state above. Pulling back to enquire into the whole of the eighteenth-century printed corpus, and designing tools fit for apprehending and making sense of such a large, impersonal, non-individual view, will result in a complementary picture of how knowledge was structured and transmitted in the eighteenth century compared with individualistic or ‘great man’ narratives of the history of ideas. Taking the eighteenth-century corpus as a field of enquiry (even in light of its various well-known limitations and inherent biases) results of course in a virtually total loss of granularity at the level of close reading, and a consequent turn away from individual texts written by individual people. But what replaces this is the ability to reconstitute structures of semantic expression from that time that are unexaminable at the level of the sentence and the paragraph, the chapter, and the book. As will be demonstrated, we may now make visible the printed, common, and possibly ‘public’ or social structures of meaning. This article’s field of enquiry is the vast common printed semantic storehouse, as opposed to what is laid down by single authors in clauses, sentences, and chapters.

BEmoC: A Corpus for Identifying Emotion in Bengali Texts.

Emotion classification in text has growing interest among NLP experts due to the enormous availability of people’s emotions and its emergence on various Web 2.0 applications/services. Emotion classification in the Bengali texts is also gradually being considered as an important task for sports, e-commerce, entertainments, and security applications. However, It is a very critical task to develop an automatic emotion classification system for low-resource languages such as, Bengali. Scarcity of resources and deficiency of benchmark corpora make the task more complicated. Thus, the development of a benchmark corpus is the prerequisite to develop an emotion classifier for Bengali texts. This paper describes the development of an emotional corpus (hereafter called ‘BEmoC’) for classifying six emotions in Bengali texts. The corpus development process consists of four key steps: data crawling, pre-processing, labelling, and verification. A total of 7000 texts are labelled into six basic emotion categories such as anger, fear, surprise, sadness, joy, and disgust, respectively. Dataset evaluation with 0.969 Cohen’s κ score indicates the close agreement between the corpus annotators and the expert. The analysis of evaluation also represents that the distribution of emotion words obeys Zipf’s law. Moreover, the results of BEmoC analysis shown in terms of coding reliability, emotion density, and most frequent emotion words, respectively.

A Quantitative Study on Dream of the Red Chamber: Word‐Length Distribution and Authorship Attribution

This paper investigates the distribution characteristics of word lengths in the Dream of the Red Chamber (DRC), measured in terms of the number of syllables or characters. The results show that the frequency distribution of words of different lengths in the DRC abides by the extended logarithmic distribution model. A comparison between the first forty, the middle forty, and the last forty chapters shows that the distribution of word lengths in these three parts does not differ significantly, which sheds light on the authorship of the novel.

Compilation, Analysis and Application of a Comprehensive Bangla Corpus KUMono

Research in Natural Language Processing (NLP) and computational linguistics highly depends on a good quality representative corpus of any specific language. Bangla is one of the most spoken languages in the world but Bangla NLP research is in its early stage of development due to the lack of quality public corpus. This article describes the detailed compilation methodology of a comprehensive monolingual Bangla corpus, KUMono. The newly developed corpus consists of more than 350 million word tokens and more than one million unique tokens from 18 major text categories of online Bangla websites. We have conducted several word-level and character-level linguistic phenomenon analyses based on empirical studies of the developed corpus. The corpus follows Zipf’s curve and hapax legomena rule. The quality of the corpus is also assessed by analyzing and comparing the inherent sparseness of the corpus with existing Bangla corpora, by analyzing the distribution of function words of the corpus and vocabulary growth rate. We have developed a Bangla article categorization application based on the KUMono corpus and received compelling results by comparing to the state-of-the-art models.

ПОЛИМОДАЛЬНЫЙ ПРОСТРАНСТВЕННЫЙ ДЕЙКСИС В РЕЧИ И ЖЕСТАХ: СИСТЕМЫ КООРДИНАТ В ЭКСПЛАНАТОРНОМ ДИСКУРСЕ

The study explores the distribution of deictic words and gestures in multimodal explanatory discourse. To proceed, we compile the corpus of speech and gesture in which the experiment participants are explaining the difference between close synonyms. The study develops the method of modelling deixis as a discourse frame of reference which allows to identify objects and situations via pointing at them. We show that two frames of reference - egocentric and allocentric - help explore the discourse semantics of deictic words and gestures on the same grounds using the coordinates of two frames which are the distance from the speaker (closer to the speaker / farther from the speaker) and the type of communicative situation in discourse (the internal communicative situation / the external communicative situation). The results reveal that the differences between the two frames of reference may not always be explicit in the semantics of deictic words, still they may appear in the use of gesture types.

Level spacing statistics for the multi-dimensional quantum harmonic oscillator: Algebraic case

We study the statistical properties of the spacings between neighboring energy levels for the multi-dimensional quantum harmonic oscillator that occur in a window [E, E + ΔE) of fixed width ΔE as E tends to infinity. This regime provides a notable exception to the Berry–Tabor conjecture from quantum chaos, and, for that reason, it was studied extensively by Berry and Tabor in their seminal paper from 1977. We focus entirely on the case that the (ratios of) frequencies ω1, ω2, …, ωd together with 1 form a basis for an algebraic number field Φ of degree d + 1, allowing us to use tools from algebraic number theory. This special case was studied by Dyson, Bleher, Bleher–Homma–Ji–Roeder–Shen, and others. Under a suitable rescaling, we prove that the distribution of spacings behaves asymptotically quasiperiodically in log E. We also prove that the distribution of ratios of neighboring spacings behaves asymptotically quasiperiodically in log E. The same holds for the distribution of finite words in the finite alphabet of rescaled spacings. Mathematically, our work is a higher dimensional version of the Steinhaus conjecture (three gap theorem) involving the fractional parts of a linear form in more than one variable, and it is of independent interest from this perspective.

BioBERT and Similar Approaches for Relation Extraction.

In biomedicine, facts about relations between entities (disease, gene, drug, etc.) are hidden in the large trove of 30million scientific publications. The curated information is proven to play an important role in various applications such as drug repurposing and precision medicine. Recently, due to the advancement in deep learning a transformer architecture named BERT (Bidirectional Encoder Representations from Transformers) has been proposed. This pretrained language model trained using the Books Corpus with 800M words and English Wikipedia with 2500M words reported state of the art results in various NLP (Natural Language Processing) tasks including relation extraction. It is a widely accepted notion that due to the word distribution shift, general domain models exhibit poor performance in information extraction tasks of the biomedical domain. Due to this, an architecture is later adapted to the biomedical domain by training the language models using 28million scientific literatures from PubMed and PubMed central. This chapter presents a protocol for relation extraction using BERT by discussing state-of-the-art for BERT versions in the biomedical domain such as BioBERT. The protocol emphasis on general BERT architecture, pretraining and fine tuning, leveraging biomedical information, and finally a knowledge graph infusion to the BERT model layer.

Modeling Deception: A Case Study of Email Phishing

Information manipulation for deception continues to evolve at a remarkable rate. Artificial intelligence has greatly reduced the burden of combing through documents for evidence of manipulation; but it has also enabled the development of clever modes of deception. In this study, we modeled deception attacks by examining phishing emails that successfully evaded detection by the Microsoft 365 filtering system. The sample population selected for this study was the University of North Texas students, faculty, staff, alumni and retirees who maintain their university email accounts. The model explains why certain individuals and organizations are selected as targets, and identifies potential counter measures and counter attacks. Over a one-year period, 432 phishing emails with different features, characters, length, context and semantics successfully passed through Microsoft Office 365 filtering system. The targeted population ranged from 18 years old up to those of retirement age; ranged across educational levels from undergraduate through doctoral levels; and ranged across races. The unstructured data was preprocessed by filtering out duplicates to avoid overemphasizing a single attack. The term frequency-inverse document frequency (TF-IDF) and distribution of words over documents (topic modeling) were analyzed. Results show that staff and students were the main target audience, and the phishing email volume spiked in the summer and holiday season. The TF-IDF analysis showed that the phishing emails could be categorized under six categories: reward, urgency, job, entertainment, fear, and curiosity. Analysis showed that attackers use information gap theory to bait email recipients to open phishing emails with no subject line or very attractive subject line in about thirty percent of cases. Ambiguity remains the main stimulus used by phishing attackers, while the reinforcements used to misinform the targets range from positive reinforcements (prize, reward) to negative reinforcements (blackmail, potential consequences).

Individual Style of Dostoevsky: Dimensions of Investigation

The article examines the existing approaches to the study of the individual style of the writer. We call these approaches models of individual style. We distinguish lexical model, syntactical model, narrative model and the model of intertextual relations. Specific features of the lexical model consist of the frequency distribution of certain words in text formation. The frequency distribution of lexical items reflects the importance of certain meanings and conceptual areas for a given author. The syntactic model of individual style is based on identifying characteristic syntactic constructions that have a high frequency rank for a particular author. The narrative model characterizes the way the story is organized. The units of the narrative model are the components of the plot and the operations that are performed on these components (moving, eliminating fragments of the narrative, combining episodes and scenes, introducing some components of the plot into others). Intertextual relations, that is, a set of semantic and formal references to other texts of the same or other authors, form an intertextual model, which also is unique for every writer. The above mentioned models complement each other and, taken together, make it possible to describe the specific features of the unique speech practices of the writer. Models of individual style are illustrated by the examples from Dostoevskys works. Modern corpus data technologies make it possible to more strictly develop the models of individual style, enhancing their objectivity by the implementation of quantitative methods. This article is a part of a project, initiated by Yu.N. Karaulov and E.L. Ginzburg.

A Continuous Semantic Embedding Method for Video Compact Representation

Video compact representation aims to obtain a representation that could reflect the kernel mode of video content and concisely describe the video. As most information in complex videos is either noisy or redundant, some researchers have instead focused on long-term video semantics. Recent video compact representation methods heavily rely on the segmentation accuracy of video semantics. In this paper, we propose a novel framework to address these challenges. Specifically, we designed a novel continuous video semantic embedding model to learn the actual distribution of video words. First, an embedding model based on the continuous bag of words method is proposed to learn the video embeddings, integrated with a well-designed discriminative negative sampling approach, which helps emphasize the convincing clips in the embedding while weakening the influence of the confusing ones. Second, an aggregated distribution pooling method is proposed to capture the semantic distribution of kernel modes in videos. Finally, our well-trained model can generate compact video representations by direct inference, which provides our model with a better generalization ability compared with those of previous methods. We performed extensive experiments on event detection and the mining of representative event parts. Experiments on TRECVID MED11 and CCV datasets demonstrated the effectiveness of our method. Our method could capture the semantic distribution of kernel modes in videos and shows powerful potential to discover and better describe complex video patterns.

Predicting standardized absolute returns using rolling-sample textual modelling

Understanding how textual information impacts financial market volatility has been one of the growing topics in financial econometric research. In this paper, we aim to examine the relationship between the volatility measure that is extracted from GARCH modelling and textual news information both publicly available and from subscription, and the performances of the two datasets are compared. We utilize a latent Dirichlet allocation method to capture the dynamic features of the textual data overtime by summarizing their statistical outputs, such as topic distributions in documents and word distributions in topics. In addition, we transform various measures representing the popularity and diversity of topics to form predictors for a rolling regression model to assess the usefulness of textual information. The proposed method captures the statistical properties of textual information over different time periods and its performance is evaluated in an out-of-sample analysis. Our results show that the topic measures are more useful for predicting our volatility proxy, the unexplained variance from the GARCH model than the simple moving average. The finding indicates that our method is helpful in extracting significant textual information to improve the prediction of stock market volatility.

“This article is interesting, however”: exploring the language use in the peer review comment of articles published in the BMJ

PurposeThe purpose of this paper is to reveal a symbol – “however” that authors are very interested in, but few research studies pay attention to the existing literature. The authors aim to further insight its function.Design/methodology/approachIn this research, the authors selected 3,329 valid comments on articles published in the British Medical Journal (BMJ) from 2015 to 2020 as the research objects. The authors showed the length distribution of reviewers' comments. In what follows, the authors analyzed the general distribution of words in comments and reviewer comments’ position to understand reviewers' comments qualitatively in word dimension. Specially, the authors analyzed functions of “however” and “but”, words that authors are most concerned with. In addition, the authors also discussed some factors, which may be related to “however,” that reflect reviewers' praise through regression analysis.FindingsThe authors found that there are marked differences in the length of reviewers' comments under different review rounds. By mapping the reviewers' comments to different sections, the authors found that reviewers are deeply concerned to methods section. Adjectives and adverbs in comments written in different sections of the manuscripts also have different characteristics. The authors tried to interpret the turning function of “however” in scientific communication. Its frequency of use is related to reviewers' identities, specifically academic status. More precisely, junior researchers use “however” in praise more frequently than senior researchers do.Research limitations/implicationsThe linguistic feature and function of “however” and “but” in the reviewers' comments of the rejected manuscripts may be different from accepted papers and also worth exploring. Regrettably, the authors cannot obtain the peer review comments of rejected manuscripts. This point may limit the conclusion of the investigation of this article.Originality/valueOverall, the survey results revealed some language features of reviewers' comments, which could provide a basis of future endeavors for many reviewers in open peer review (OPR) field. Specially, the authors also put forward an interesting symbol to examine the review comments, “however”, for the first time.

Online Latent Dirichlet Allocation Model Based on Sentiment Polarity Time Series

The Product Sensitive Online Dirichlet Allocation model (PSOLDA) proposed in this paper mainly uses the sentiment polarity of topic words in the review text to improve the accuracy of topic evolution. First, we use Latent Dirichlet Allocation (LDA) to obtain the distribution of topic words in the current time window. Second, the word2vec word vector is used as auxiliary information to determine the sentiment polarity and obtain the sentiment polarity distribution of the current topic. Finally, the sentiment polarity changes of the topics in the previous and next time window are mapped to the sentiment factors, and the distribution of topic words in the next time window is controlled through them. The experimental results show that the PSOLDA model decreases the probability distribution by 0.160 1, while Online Twitter LDA only increases by 0.069 9. The topic evolution method that integrates the sentimental information of topic words proposed in this paper is better than the traditional model.

Modeling Global and Local Interactions for Online Conversation Recommendation

The popularity of social media platforms results in a huge volume of online conversations produced every day. To help users better engage in online conversations, this article presents a novel framework to automatically recommend conversations to users based on what they said and how they behaved in their chatting histories. While prior work mostly focuses on post-level recommendation, we aim to explore conversation context and model the interaction patterns therein. Furthermore, to characterize personal interests from interleaving user interactions, we learn (1) global interactions , represented by topic and discourse word clusters to reflect users’ content and pragmatic preferences, and (2) local interactions , encoding replying relations and chronological order of conversation turns to characterize users’ prior behavior. Built on collaborative filtering, our model captures global interactions via discovering word distributions to represent users’ topical interests and discourse behaviors, while local interactions are explored with graph-structured networks exploiting both reply structure and temporal features. Extensive experiments on three datasets from Twitter and Reddit show that our model coupling global and local interactions significantly outperforms the state-of-the-art model. Further analyses show that our model is able to capture meaningful features from global and local interactions, which results in its superior performance in conversation recommendation.

From Keyness to Distinctiveness – Triangulation and Evaluationin Computational Literary Studies

From Keyness to Distinctiveness – Triangulation and Evaluation<b>in Computational Literary Studies</b>

MXene-enhanced β-phase crystallization in ferroelectric porous composites for highly-sensitive dynamic force sensors

Piezoelectric polyvinylidene fluoride (PVDF) has been widely utilized in flexible and self-powered tactile sensors, which require high ferroelectricity of polar phase PVDF. Herein, we demonstrate self-powered piezoelectric e-skins with high sensitivity and broad sensing range based on 3D porous structures of MXene (Ti 3 C 2 T x )/PVDF. MXene was used as a nucleation agent to increase the ferroelectric properties of PVDF. This was carried out considering its 2D geometry and abundant surface functional groups that facilitate intermolecular hydrogen bonding between the surface functional groups of MXene and the CH 2 group of PVDF. In addition, porous structures can increase the variation in contact area and localized stress concentration in response to applied pressure. This further enhances the piezoelectric sensitivity. Owing to structural deformation and localized stress concentration, the piezoelectric sensitivity of porous MXene/PVDF e-skin is 11.9 and 1.4 nA kPa −1 for low (< 2.5 kPa) and high (2.5–100 kPa) pressure ranges, respectively. These are 31 and 3.7 times higher, respectively, than that of planar MXene/PVDF e-skin (0.4 nA kPa −1 for <100 kPa). In addition, porous MXene/PVDF e-skin exhibits a broad sensing range of up to 100 kPa, and stable sensing performance (5000 repetitions). Our piezoelectric porous MXene/PVDF e-skins enable the monitoring of high-frequency dynamic signals such as acoustic sound waves as well as low-frequency radial artery pulses. In particular, the detection of high-frequency vibrations from sliding friction enables our sensor array to perceive various surface textures with different roughness and moduli, as well as the spatial distribution of words embossed on surfaces. This demonstrates its substantial potential for application in wearable devices, prosthetic limbs, robotics, and healthcare monitoring devices. MXene with a large specific surface area and abundant functional groups can be used as a heterogeneous nucleation agent to induce crystallization of the polar β-phase and enhance the ferroelectric properties of PVDF, which enables the monitoring of high-frequency dynamic signals such as acoustic sound waves as well as low-frequency radial artery pulses. In particular, the detection of high-frequency vibrations from sliding friction enables our sensor array to perceive various surface textures with different roughness and moduli, as well as the spatial distribution of words embossed on surfaces. • High-performance piezoelectric skin sensor is demonstrated using 3D porous structures of MXene (Ti 3 C 2 T x )/PVDF composites. • MXene is used as a nucleation agent to induce polar β-phase and enhance the ferroelectric properties of PVDF. • MXene/PVDF sensor perceives various surface textures as well as the spatial distribution of words embossed on surfaces.

Analysis Of Pedagogical Texts Prepared For Academic Turkish As A Foreign Language In Terms Of Academic Vocabulary

In this study, the aim was to determine academic words in academic Turkish teaching sets, to describe the numerical and rational distribution of these academic words and to make a comparison between the vocabulary of engineering-health sciences and social sciences. Vocabulary units classified as social sciences, and engineering-health sciences were evaluated in terms of their percentages in their own categories, the most frequently used word types and conceptual features and to analysed the differences in usage between social and engineering-health science, Mann Whitney U test was adapted. The result of the study indicated that frequency of academic words’ usage (%7-%8) is lower than their use count; however, when the number of single uses were considered, in terms of developing comprehension and explanation skills, it is highly important. When social science and engineering-health science vocabulary were compared in terms of vocabulary frequency and academic word’s frequency.